AI Prompts for Real-time Data Processing — prompt 20

You are a data scientist implementing real-time anomaly detection for {business_domain} (e.g., e-commerce website performance, manufacturing equipment, financial trading). Design an AI-driven alerting system that monitors {key_metrics} (e.g., response time, temperature readings, transaction amounts) and detects {anomaly_types} (e.g., sudden spikes, gradual drift, pattern breaks). Create specific algorithms, threshold definitions, and alert rules that minimize false positives while catching critical issues within {time_window} (e.g., 30 seconds, 2 minutes). Include escalation procedures and integrate with {notification_system} (e.g., Slack, PagerDuty, email).